Optimization of Wireless Positioning and Sensing Network (WPSN) Architecture using Machine Learning Technique

点击翻译按钮获取中文摘要

Wireless Positioning and Sensing technology is emerging as the latest technological strategy in tracking objects using magnetic or electromagnetic response exchange. Wireless readers that are key components of this technology are equipped with antennas for sending and receiving signals, a transceiver and a processor to decode data. Finding optimal number and placement of these readers for indoor and outdoor purposes are an important issue in these days.Recently, Peytec Inc, designed and fabricated smart sensors for Wireless Positioning and Sensing Network (WPSN) application using scalable wireless communication system to sense and track oudoor/indoor object location data to create automated transactions to the inventory system. Sensing location information in indoor industrial environments requires a high precision wireless communication system and is a challenging task in part because various objects reflect and disperse signals which adversely affects reliability of communication of sensing and location data. Currently, Peytec Inc. is only capable of collecting the information when no obstacle is present (within an acceptable accuracy range of 10 cm), however once an obstacle is introduced the accuracy drops dramatically as the travel time changes.The objective of this project is to develop machine learning algorithms to localize object with required precision, and to architect optimum Wireless Positioning and Sensing Network (WPSN) with respect to existing static and dynamic obstacles for indoor localization purposes. This project will pull together talents from Ryerson University and Peytec Inc. and take the advantage of the resources of Ryerson including experienced researchers, energetic and goal-driven graduate students with a track record of excellence of Peytec Inc. in design and development of WPSN to jointly explore the optimal number of required readers for tracking products by utilizing historical data, training data and in particular, Machine Learning (ML) approaches. Successful design of optimization problems and ML algorithms to identify optimum WPSN architectures will enable new differentiated product capabilities, market expansion and sales for Peytec Inc. Ontario, Canada. In addition, this project will provide a rich environment for the training of HQP by having them engaged in development and design of industry-caliber products.